Vals AI Updates

How Far Will AI Assistants Go for a Refund They Can’t Get?

Routers Are a New AI Primitive

Introducing Robo-bench: can a model control a robot?

Meta's Wonder and Wonder Full evaluated across our benchmark suite

zAI's GLM Legal evaluated on our legal benchmarks

Do Agent Teams Pay Off? A Case Study on Vibe Code Bench

Anthropic's Claude Haiku 5.5 evaluated across our benchmark suite

Introducing SAFE-Teen: evaluating AI chatbots on teen mental health

An Independent Evaluation of TypeSafe's Jev

Claude Opus 5.5 agents find two room-temperature magnetic semiconductor candidates